Development of a Potential-Modulated Electrochemiluminescence Measurement System for Selective and Sensitive Determination of the Controlled Drug Codeine

Abstract

Codeine is a common analgesic drug that is a pro-drug of morphine. It also has a high risk of abuse as a recreational drug because of its extensive distribution as an OTC drug. Therefore, sensitive and selective screening methods for codeine are crucial in forensic analytical chemistry. To date, a commercial analytical kit has not been developed for dedicated codeine determination, and there is a need for an analytical method to quantify codeine in the field. In the present work, potential modulation was combined with electrochemiluminescence (ECL) for sensitive determination of codeine. The potential modulated technique involved applying a signal to electrodes by superimposing an AC potential on the DC potential. When tris(2,2′-bipyridine)ruthenium(II) ([Ru(bpy)₃]²⁺) was used as an ECL emitter, ECL activity was confirmed for codeine. A detailed investigation of the electrochemical reaction mechanism suggested a characteristic ECL reaction mechanism involving electrochemical oxidation of the opioid framework. Besides the usual ECL reaction derived from the amine framework, selective detection of codeine was possible under the measurement conditions, with clear luminescence observed in an acidic solution. The sensitivity of codeine detection by potential modulated-ECL was one order of magnitude higher than that obtained with the conventional potential sweep method. The proposed method was applied to codeine determination in actual prescription medications and OTC drug samples. Codeine was selectively determined from other compounds in medications and showed good linearity with a low detection limit (150 ng mL⁻¹).

Introduction

Codeine is a widely recognized pro-drug of morphine, wherein the analgesic effect necessitates hepatic metabolism through O-demethylation to yield morphine^1,2) (Fig. 1). Codeine is a low-potency opioid utilized for the management of mild to moderate pain, and it is classified as the intermediate phase in the triadic analgesic hierarchy.³⁾ Its advantageous impact assumes paramount significance in the clinical domain, and codeine stands out as one of the most commonly employed analgesics, so the quantities of codeine produced and consumed have exceeded 200 t annually on a global scale since 2007.⁴⁾ In the in vivo setting, the unaltered form of codeine exhibits antitussive effects,⁵⁾ and it is disseminated as a constituent in non-prescription medications. Regrettably, individuals with hyperactive genetic CYP2D6 activity encounter elevated concentrations of the metabolite morphine in their bloodstream, leading to a heightened likelihood of respiratory depression and documented instances of fatality.⁶⁾ Furthermore, there exists a suggestion that codeine has the potential to engender manifestations of addiction, encompassing pharmacological reliance accompanied by feelings of euphoria.⁷⁾ Due to its widespread accessibility and susceptibility to misuse, codeine has historically posed a significant societal concern. Codeine is a substance with considerable potential for misuse and is subject to regulatory measures as a narcotic or psychotropic agent in several nations and jurisdictions. Regulatory prerequisites vary based on the specified concentration of codeine delineated in the legislation and regulatory frameworks of each individual nation and jurisdiction. Nevertheless, owing to the ongoing clinical utility of codeine, its governance is comparatively lenient for non-prescription medications that incorporate minimal concentrations of the compound. Henceforth, the identification and quantification of codeine in authentic specimens, including commercially available medications and illicit substances retrieved from real criminal settings, should imperative not only for qualitative detection in clinical diagnostics but also for the advancement of forensic analytical chemistry in on-site inquiries conducted at crime scenes.⁸⁾

Fig. 1. Chemical Structures of Codeine and Morphine

The utilization of mass spectrometry (MS), coupled with either gas chromatography or liquid chromatography (LC), represents the foremost preference in the process of codeine determination due to its capability to offer exceedingly elevated levels of sensitivity and selectivity.^9,10) Nevertheless, these methodologies necessitate comparatively intricate pretreatment procedures, substantial and cumbersome apparatus, and exhibit elevated analytical expenses. Additionally, gas chromatography-MS and LC-MS analyses possess relatively protracted durations, thereby impeding their on-site applicability in scenarios that demand prompt outcomes and the screening of numerous specimens, such as point-of-care and roadside drug tests. Consequently, colorimetric detection techniques, encompassing immunoassays, have recurrently been employed for straightforward screening purposes.^11–13) Nonetheless, the colorimetric approach is frequently conjoined with enzymatic or antigen–antibody reactions. Consequently, meticulous storage of the analysis kits is essential, and the associated analytical expenses often exhibit a relatively elevated magnitude. Moreover, owing to the relatively diminished selectivity exhibited by these colorimetric methods, their suitability for codeine-specific determinations is inadequate.^14–16) Furthermore, employing alterations in color intensity as a basis for quantitative analysis in colorimetric methods proves to be highly challenging. Therefore, there is a pressing need for a quantitative analytical approach that demonstrates selectivity towards codeine and allows for on-site analysis.

Electrochemical measurement represents a portable methodology that has been employed as a straightforward and highly sensitive determination method for on-site analytical systems.¹⁷⁾ Numerous electrochemical methodologies have been documented for the determination of codeine. Well-designed modified electrodes have prepared for sensitive determination of codeine in actual samples.^18–22) However, due to the meticulous precision demanded during the fabrication of modified electrodes, the utilization of specialized systems for these electrodes is frequently imperative, alongside the indispensability of precise expertise in their production.

Electrochemiluminescence (ECL) denotes a phenomenon whereby species generated at electrodes undergo electron transfer reactions, resulting in the formation of an excited state that emits light.^23–30) The ECL can be generated utilizing electrochemical apparatus, and this technique has garnered substantial attention in the field of analytical chemistry for its application in on-site determinations. The ECL reaction with the typical ECL emitter tris(2,2′-bipyridine)ruthenium (II) complex ([Ru(bpy)₃]²⁺) has high sensitivity for aliphatic amines.^26,31) Some literature have exhibited the application of ECL measurements to determine codeine.^32,33) However, since the nature of ECL being generated through the formation of an excited ECL emitter, its inherent selectivity is essentially low. The ECL reaction is significantly influenced by the occurrence of false positives from various amines such as ephedrine, diphenhydramine, and chlorpheniramine, which are frequently present in OTC medications. Furthermore, the electrochemical oxidation activity of codeine is relatively low, posing challenges for the selective and quantitative detection of codeine. To the best of our knowledge, the elucidation of the ECL reaction mechanism involving the [Ru(bpy)₃]²⁺/codeine system remains elusive, and the optimization of ECL measurement conditions for this system is yet to be accomplished. The aim of this study is to establish an on-site ECL measurement protocol for the selectively detection of codeine in OTC drugs containing interfering amines. Furthermore, we explored the applicability of our highly sensitive codeine quantification method for on-site analysis in practical settings. In general, since the electrochemical oxidation kinetics of amines is relatively low. Therefore, the amount of radical intermediates produced during the electrochemical oxidation process of codeine is low, the absolute ECL is essentially weak. Thus, to enhance the sensitivity and signal-to-noise (S/N) ratio of the ECL signal, we employed a potential modulation (PM) technique during the ECL measurement.^34–37) Source modulation can provide substantial enhancements in the detection limits of analysis by virtue of diminishing background signals. In contrast to conventional potential-modulated ac voltammetry, PM-ECL offers the advantage of acquiring analytical signals through larger amplitude sine wave voltage perturbations, which augmenting the analytical sensitivity.

Experimental

Reagents and Materials

Tris(2,2′-bipyridine)ruthenium(II) chloride [Ru(bpy)₃]Cl₂ hexahydrate was purchased from Sigma-Aldrich (St. Louis, MO, U.S.A.) and used without further purification. Codeine phosphate and morphine hydrochloride were obtained from Takeda Pharmaceutical Co. Ltd. (Tokyo, Japan). DL-Methylephedrine was purchased from Alps Pharmaceutical Ind. Co., Ltd. (Gifu, Japan). Chlorpheniramine maleate and ascorbic acid were purchased from Wako Pure Chemical Corporation (Osaka, Japan). Working standard solutions were prepared by precise dilution of stock solutions with water. Phosphate-buffered saline (PBS) was prepared by mixing disodium hydrogen phosphate and potassium dihydrogen solutions with known volumes and concentrations (Tokyo Chemical Industry Co., Ltd., Tokyo, Japan). The pH was adjusted with phosphoric acid or an aqueous sodium hydroxide solution. All other reagents were of analytical grade and purchased from Nacalai Tesque Inc. (Kyoto, Japan). Ultrapure water (resistivity >18.2 MΩ·cm) was prepared using a Milli-Q water purification system (Millipore, Billerica, MA, U.S.A.).

Apparatus

A glassy carbon electrode (ø 3.0 mm), a Ag/AgCl electrode (saturated KCl), and platinum wire were used as the working, reference, and counter electrodes, respectively. Before each experiment, the working electrode was polished with 0.30-µm aluminum powder on a diamond pad and sonicated in Milli-Q water using a Brasonic ultrasonic cleaner (40 kHz, 160 W; Branson Ultrasonics, Danbury, CT, U.S.A.). The electrochemical cell was placed in a light-proof box. The electrode potential was controlled by a type 2000C potentiostat/galvanostat (TOYO Corporation, Tokyo, Japan) combined with a HECS980 Potential Sweep Unit (Fuso, Kanagawa, Japan). An H11901-01 photomultiplier tube module equipped with a C7319 signal amplifier unit (Hamamatsu Photonics, Shizuoka, Japan) was placed 1.0 mm from the working electrode to detect the ECL. A C7169 power supply unit (Hamamatsu Photonics) was used to drive the photosensor module.

Supplementary Figure S1 shows a schematic diagram of the experimental setup for the PM-ECL measurements. A sinusoidal AC voltage generated by a function generator (model AD-8624A, A&D Co. Ltd., Tokyo, Japan) was superimposed on a DC potential ramp during the potential scan. The output ECL signal from the photomultiplier tube photosensor module was amplified by a 5101 lock-in-amplifier (EG&G/PAR, MA, U.S.A.) and the current signal from the potentiostat was fed into an AIO-160802AY-USB AD/DA converter (Contec Co., Ltd., Osaka, Japan).

To validate the proposed PM-ECL method, LC-QTOF MS analyses were performed using a Nexera liquid chromatograph (Shimadzu Co., Kyoto, Japan) combined with a LCMS-9030 mass spectrometer (Shimadzu Co.). A Kinetex^® XB-C18 column (100 mm × 2.1 mm i.d.; 2.6 µm particle size; Phenomenex, Torrence, CA, U.S.A.) was used for separation and kept at 40 °C during the analysis. The mobile phase was a mixture of 0.1% (v/v) formic acid–10 mM ammonium formate (solvent A) and 0.1% (v/v) formic acid–10 mM ammonium formate–methanol (solvent B). Both components of the mobile phase were filtered and degassed before use. The elution used a linear gradient of solvent B from 5 to 95% (7.5 min), with a constant mobile phase flow rate of 0.3 mL min⁻¹. The mass spectra were obtained in full scan mode. The ion at m/z 300.1594 ± 0.0005 ppm (electrospray ionization positive) was monitored for codeine.

Sample Preparation

Prescription drugs (sample A [tablet] and B [powder]) and OTC combination cold remedies (sample C [tablet], and D [medicated syrup]) containing codeine were purchased at a pharmacy. Samples A, B, and C were homogenized in a mortar, and 0.431 g of each powder was dispersed in 50 mL of Milli-Q water in a separate test tube by ultrasonication. Sample D was directly diluted 10 times with Milli-Q water in the test tube. The test tubes were centrifuged in an H-19α centrifuge (Kokusan Co. Ltd., Saitama, Japan) for 10 min at 2650 × g. The supernatant from each sample solution was collected from the test tube and 100 mL was transferred to the ECL measurement cell. PBS (900 µL, 0.1 M) containing 500 µM of [Ru(bpy)₃]²⁺ was added to each electrochemical cell, the PM-ECL measurement was carried out.

Results and Discussion

Electrochemistry and ECL Behavior of the [Ru(bpy)₃]²⁺/Codeine System

Although ECL of [Ru(bpy)₃]²⁺/codeine system has been reported,^32,38) the detailed reaction mechanism has not been elucidated. First, we characterized the ECL reaction of the [Ru(bpy)₃]²⁺/codeine system. Cyclic voltammetry (CV; Fig. 2A) and ECL signals (Fig. 2B) were obtained for 500 µM [Ru(bpy)₃]²⁺ in 0.1 M PBS (pH 5.0) under conventional potential sweep. The dashed lines show the background responses. A redox peak pair (midpoint potential) of [Ru(bpy)₃]²⁺/[Ru(bpy)₃]³⁺ was observed at +1.1 V (vs. Ag/AgCl) in the CV results (Fig. 2A, dashed line). When 500 µM codeine was introduced to the solution, a weak oxidation current appeared from +0.95 V. The electrode potential was further shifted to the positive direction, and an ECL peak was observed close to the oxidation potential of [Ru(bpy)₃]²⁺ (Fig. 2B, blue line). ECL emits light when [Ru(bpy)₃]³⁺ produced by electrode reactions is reduced by radical intermediates of the co-reactant, resulting in an excited state of [Ru(bpy)₃]²⁺*.

Fig. 2. (A) Cyclic Voltammograms (CVs) and (B) Corresponding Electrochemiluminescence (ECL) Profiles for 500 µM [Ru(bpy)₃]²⁺ in Phosphate-Buffered Saline (PBS, pH 5.0) in the Presence of 500 µM Codeine

The dashed lines show the responses for the 500 µM [Ru(bpy)₃]²⁺ background.

The generation process and stability (lifetime) of the radical intermediates are affected by the pH, since the deprotonation rate was essentially dependent on the pH. We investigated the pH dependence of the ECL intensity. Figure 3a shows the dependence the ECL intensity for the [Ru(bpy)₃]²⁺/codeine system on pH. The [Ru(bpy)₃]²⁺/codeine system showed an ECL response under weak basic conditions (> pH 9). Codeine has a tertiary amine framework in its structure, and should undergo an ECL mechanism similar to that of tripropylamine, a typical co-reactant (Supplementary Chart S1). The pK_a of codeine is reportedly 8.2,³⁹⁾ and the free-base form is dominant in weakly alkaline solutions. The strongly reducing radical species of codeine (codeine·) is believed to be formed by loss of an electron and a proton from the α-carbon of the amine site in an electrode reaction. The excited state of [Ru(bpy)₃]²⁺ was generated through electron transfer between [Ru(bpy)₃]³⁺ and codeine. The ECL intensity of the codeine system was similar to that of the morphine system (Fig. 3b). Interestingly, clear ECL was also observed under weakly acidic conditions for the codeine system, while a negligible ECL signal was observed for the morphine system. The ECL intensity of the codeine system at pH values below 7 was higher than that under weakly alkaline conditions, which suggested a different ECL route.

Fig. 3. Dependence of pH on the Electrochemiluminescence (ECL) Intensity for 500 µM Ru(bpy)₃²⁺ in the Presence of (a) 500 µM Codeine (Red Circles) and (b) 500 µM Morphine (Blue Circles)

(c) The results in the absence of co-reactant (background).

Under weakly acidic conditions, codeine was oxidized at approximately +1.0 V in squarewave voltammetry (Supplementary Fig. S2). The oxidation potential shifted to the negative direction with increasing solution pH and showed a Nernst response. The results suggested that the electrode reaction of codeine under weakly acidic conditions involved protons other than those on the amine framework. The electrochemical oxidation mechanism of codeine has not been thoroughly analyzed in detail, and the electrochemical oxidation intermediates have not been elucidated.⁴⁰⁾ Morphine undergoes an electrochemical oxidation process to form the dimeric form pseudomorphine (Eq. 1; R = H).

  

Supplementary Fig. S3 shows the results of density functional theory calculations for the electron density distribution of morphine. The highest occupied molecular orbital orbitals of morphine show that the electron density coefficient is localized on the phenol framework, confirming its high reactivity. Therefore, radicals generated during the electrochemical oxidation process of the phenol skeleton were rapidly consumed by the coupling reaction. This reaction is relatively fast and a distinct oxidation current is observed.⁴¹⁾ Therefore the morphine radical intermediates produced by the electrode reaction rapidly disappeared and no ECL response was observed under weakly acidic conditions. Besides, anisole radicals are more stable than phenoxy radicals and remain in solution.⁴²⁾ Furthermore, for codeine (R = CH₃), the methoxy group at position 3 of the morphine framework causes steric hindrance in the dimerization reaction. Therefore, the lifetime of the codeine radical, which is an anisole radical, is prolonged. As a result, the radical intermediates on the morphine skeleton of codeine generated by electrochemical oxidation in the vicinity of the electrode surface. On the basis of the above discussion of the electrochemical oxidation process of codeine, the ECL mechanism of the codeine/[Ru(bpy)₃]²⁺ system in an acidic solution was assumed to be as shown in Chart 1. Codeine radical (codeine·) was generated by the oxidation process on the morphine framework of codeine (Eq. 2). Simultaneously, [Ru(bpy)₃]³⁺ was generated by the electrode reaction (Eq. 3). An excited state of [Ru(bpy)₃]²⁺* was generated by the subsequent chemical reaction with codeine· and [Ru(bpy)₃]³⁺ (Eq. 4). Consequently, ECL was observed from the electrode surface (Eq. 5).

Chart 1

According to the proposed ECL reaction mechanism of the [Ru(bpy)₃]²⁺/codeine system, the ECL intensity is proportional to the concentration of codeine. Therefore, the dependence of ECL intensity on the codeine concentration was investigated (Supplementary Fig. S4). The ECL intensity increased linearly with codeine concentration up to 200 µM, with a detection limit of approximately 6.5 µM (1.9 µg/mL). For determining codeine, the sensitivity obtained may not be sufficient when dealing with trace samples in clinical and forensic analysis. Additionally, because the ability to detect ECL is strongly influenced by the performance of the optical detector, a technique to increase the detection sensitivity is needed.

PM-ECL Behavior of the [Ru(bpy)₃]²⁺/Codeine System

We previously developed a sensitive ECL analysis method using the PM technique,^37,43) and applied PM-ECL to the sensitive determination of codeine. An alternating potential is superimposed on a direct current, and the alternating component of the observed ECL response is detected under PM-ECL measurement. Figure 4A,a shows the image of the applied potential and Fig. 4A,b shows the ECL response obtained under the conventional potential sweep method. With 20 µM codeine, the ECL signal was too weak compared with the relatively high noise. Figure 4B,a shows the image of the applied potential and Fig. 4B,b shows the detected PM-ECL response when a 100 mV amplitude and 100 Hz AC potential were superimposed on the DC ramp (20 mV/s). A clear signal was detected for 20 µM codeine using PM-ECL. The PM-ECL measurement could extract and detect the ECL signal corresponding to the AC component, which reduced the influence of noise. In addition, the PM-ECL signal introduced into the lock-in amplifier was detected by amplifying the response of the 100 Hz AC component of the reference signal. Because the PM-ECL method could be used as a high-sensitivity codeine detection method, we studied the PM-ECL response in detail and optimized the conditions.

Fig. 4. (a) Conceptual Applied Potential Profile of (A) Conventional ECL Measurement (Potential Sweep) and (B) PM-ECL Measurement

(A, b) ECL profile and (B, b) PM-ECL profile for 20 µM codeine in PBS (pH 5.0) containing 500 µM [Ru(bpy)₃]²⁺. The dashed lines are the responses for the 500 µM [Ru(bpy)₃]²⁺ background. The DC potential scan rate was 20 mV s⁻¹, the AC amplitude was 100 mV and the AC frequency was 100 Hz. The PMT-biased voltage was 800 V.

Figure 5 shows the dependence of PM-ECL on the AC amplitude for the [Ru(bpy)₃]²⁺/codeine system. The PM-ECL intensity increased with the AC amplitude. Generally, conventional AC voltammetry has been performed with a small AC amplitude, such as several millivolts. Large amplitudes are difficult to measure with AC voltammetry because the charging current interferes with the electrochemical behavior.⁴⁴⁾ By contrast, a large amplitude can be applied for PM-ECL measurements because the light emission is produced by a homogeneous chemical reaction that generates [Ru(bpy)₃]²⁺*. Therefore, the sensitivity for codeine could be increased by the increasing the amplitude. However, the PM-ECL curves broadened with increasing AC amplitude. The PM method extracts and detects the AC component. Therefore, the magnitude of the change in the ECL response during a single cycle does not alter even if the amplitude is too high. When amplitudes greater than 100 mV were used, the sensitizing effect was not as apparent even when the amplitude was increased. Considering the reproducibility of the measurement, the optimum AC amplitude for highly sensitive measurement of codeine was 100 mV.

Fig. 5. Dependence of the PM-ECL Intensity on the AC Amplitude at +1.1 V for 500 µM Ru(bpy)₃²⁺ in the Presence of (a) 20 µM Codeine and (b) Absence of Codeine

The DC potential scan rate was 20 mV s⁻¹, and the AC frequency was 100 Hz. The PMT-biased voltage was 800 V.

Figure 6 shows the dependence of PM-ECL intensities on the AC frequency. A pronounced PM-ECL signal was obtained when the modulation frequency was varied from 5 to 100 Hz. However, the PM-ECL intensity decreased when the frequency was higher than 100 Hz. ECL is a luminescence phenomenon caused by subsequent chemical reactions between chemical species generated by electrode reactions (Eq. 4). The PM measurement depends on the AC frequency when the chemical reaction rate is a rate-determining step. Figure 7 shows the raw responses for the current and ECL in the PM-ECL measurement. Figures 7A and B show the corresponding current and ECL responses in the PM-ECL measurement. The complex current and an ECL response corresponding to the potential modulation were observed periodically. To analyze the behavior of [Ru(bpy)₃]²⁺ and codeine in detail when modulated potentials were applied, magnified views of the current and ECL around the redox potential were obtained (Figs. 7C, E). Figure 7D shows a magnified view of the applied potential. The current response showed a similar waveform to the applied potential. The electrode response is relatively fast, and the current response corresponds to the change in electrode potential. The ECL response was observed with a delay to the electrode potential waveform. The subsequent chemical reaction rate in the ECL reaction was relatively slow and did not follow the modulated potential. Therefore, the PM-ECL intensity decreased at high AC frequencies. Furthermore, at higher frequencies, [Ru(bpy)₃]³⁺ generated by electrochemical oxidation is rapidly reduced to [Ru(bpy)₃]²⁺ by an electrode reaction. Generally, the stability of radical intermediates generated during the electrochemical oxidation process is relatively low under aqueous conditions. Therefore, the codeine radical at the electrode surface was deactivated before the subsequent chemical reaction with [Ru(bpy)₃]³⁺. Because of these two reasons, the PM-ECL intensity decreased at higher frequencies, and the optimum frequency was set at 100 Hz for the subsequent experiments.

Fig. 6. Dependence of the PM-ECL Intensity on the AC Frequency at +1.1 V for 500 µM [Ru(bpy)₃]²⁺ in the Presence of (a) 20 µM Codeine and (b) Absence of Codeine

The DC potential scan rate was 20 mV s⁻¹, and the AC amplitude was 100 mV. The PMT-biased voltage was 800 V.

Fig. 7. (A) Current–Time Profile and (B) Corresponding ECL–Time Profile under PM Measurement for 500 µM [Ru(bpy)₃]²⁺ and 500 µM Codeine

Magnified views of the current and ECL responses are shown in (C) and (E), respectively. (D) The actual modulation potential applied to the electrode. The DC potential scan rate was 20 mV s⁻¹, the applied frequency was 100 Hz, and the AC amplitude was 100 mV. The PMT-biased voltage was 800 V.

Analytical Performance of PM-ECL Measurements for Codeine Determination

PM-ECL measurements were carried out for 500 µM [Ru(bpy)₃]²⁺ in PBS (pH 5.0) with different concentrations of codeine (Fig. 8). The inset shows the calibration curve. A linear relationship was observed between 0 and 30 µM with a good determination coefficient (r² = 0.991). The codeine detection limit was approximately 0.50 µM (150 ng/mL). The determination range could be adjusted by manipulating the sensitivity of the photodetector, and it was possible to quantify codeine at concentrations up to approximately 1 mM. The sensitivity of the PM-ECL measurement was more than an order of magnitude higher than the conventional potential sweep method.

Fig. 8. Dependence of the PM-ECL Profile on the Concentration of Codeine

The inset shows the PM-ECL intensity at +1.1 V as a function of the concentration of codeine. The DC potential scan rate was 20 mV s⁻¹, the applied frequency was 100 Hz, and the AC amplitude was 100 mV. The PMT-biased voltage was 800 V.

Simultaneously, the selectivity of codeine for PM-ECL measurements was investigated. Some chemical species that contain an amine framework, such as ephedrine and chlorpheniramine, exhibit ECL activity with [Ru(bpy)₃]²⁺. It was difficult to selectively detect codeine under the weakly alkaline conditions frequently used for ECL measurements (Supplementary Fig. S5). Therefore, solvent extraction and solid phase extraction technique were required for selective detection.^45,46) In this study, a reaction mechanism was proposed in which codeine exhibited ECL under acidic conditions. The results of PM-ECL measurement at pH 5.0 are shown in Fig. 9. The red bars show the results for PM-ECL measurements in solutions containing codeine and interfering components. The black bars show the results for PM-ECL measurements in solutions without codeine. Codeine was selectively detected in the solution even in the presence of pharmacologically active interfering components (morphine, ephedrine, acetaminophen, and chlorpheniramine). Ascorbic acid has antioxidant capacity and ECL activity under weakly alkaline conditions. It has been widely used as an antioxidant and can deactivate the excited state of [Ru(bpy)₃]²⁺. In our previous work, we reported that a self-quenching effect occurred in the Stern–Volmer relationship through an electron exchange reaction.^47–49) The frequency of collisions between chemical species affects this relationship. The deactivation of [Ru(bpy)₃]²⁺* was dominant with high concentrations (millimolar) of ascorbic acid. Additionally, it has been confirmed that the self-quenching reaction is negligible at concentrations of 1 mM or less. On the other hand, Ru(bpy)₃³⁺ formed by electrochemical oxidation can also undergo an electron transfer reaction with ascorbic acid. Since the electron transfer is a homogeneous reaction in the solution in the vicinity of the electrode, the effect is suppressed under low concentrations of ascorbic acid. In our previous study, the ECL reaction showed a linear relationship at ascorbic acid concentrations below 1.0 mM.⁴⁷⁾ By sufficiently diluting the sample, codeine could be determined without interference from ascorbic acid. The dilution process increased the selectivity of codeine detection but decreased the absolute detection level of codeine in the measurement solution. Therefore, the highly sensitive detection method using PM-ECL is helpful for selective detection. However, several OTC drugs included morphine-like components (e.g., dextromethorphan, dimemorphan, and dihydrocodeine) as pharmacologically active species. We are currently conducting a detailed study of the ECL reaction under the weak acid conditions reported in this paper for selective determination of the individual components.

Fig. 9. Codeine Selectivity Results Using PM-ECL Measurements

Red bars show PM-ECL intensities for solutions containing 50 µM codeine and 50 µM interfering components in pH 5.0 PBS. Black bars indicate the PM-ECL response to the 50 µM interfering component only. The conditions for PM-ECL were the same as in Fig. 8. Abbreviations: ME, methylephedrine; AP, acetaminophen; AA, ascorbic acid; and CP, chlorpheniramine.

Selective and Sensitive Detection of Codeine in OTC Drugs Containing Interfering Ingredients

PM-ECL measurements were used to determine codeine in OTC medications. The samples were diluted in pure water and mixed with a pH 5 buffer solution containing [Ru(bpy)₃]²⁺ for the PM-ECL measurements. Table 1 shows the results of the PM-ECL measurements for the actual samples. The solid samples, tablets (Sample A), and powder (Sample B) contained only codeine, which has antitussive activity as a pharmacological component. The PM-ECL results were close to the declared concentrations of codeine in the products. Samples C and D, which contained methylephedrine and chlorpheniramine that could interfere with the ECL measurement, were also analyzed by PM-ECL. Although active interfering components were contained in the samples, the detected codeine concentrations were close to the declared concentrations and LC-MS results. Sample D was a highly viscous and turbid syrup, and the PM-ECL results were close to the indicated values with good reproducibility (relative standard deviation: 0.15%). All results agreed with those obtained by LC-MS.

Table 1. Results for the Determination of Codeine in Drugs and Medicines by PM-ECL

Sample	Indicated value	Found (PM-ECL)	Detection rate (%)	RSD** (%)	LC-MS results
Sample A*a) (tablet)	4.5 (mg/tablet)	4.2 (mg/tablet)	93	0.080	4.3 (mg/tablet)
Sample B*b) (powder)	4.7 (mg/g)	4.4 (mg/g)	94	0.13	4.2 (mg/g)
Sample C*c) (tablet)	4.2 (mg/tablet)	4.3 (mg/tablet)	102	0.23	4.4 (mg/tablet)
Sample D*d) (medicated syrup)	0.83 (mg/mL)	0.79 (mg/mL)	95	0.15	0.78 (mg/mL)

* Components: a) lactose hydrate, hydroxypropylcellulose, and magnesium stearate; b) lactose hydrate; c) methylephedrine (ME), chlorpheniramine (CP) maleate, anhydrous caffeine, and lactose; d) ME, CP maleate, and anhydrous caffeine. ** RSD: Relative standard deviation.

To test the proposed method for quantitative analysis of codeine, a spiking and recovery test was performed on dilute solutions of the samples (Supplementary Table S2). Good recovery rates for codeine were obtained for dilute solutions of samples A–D. An immunoassay kit used for on-site analysis was also applied. The sensitivity of the PM-ECL assay was comparable to that of commercially available immunoassay kits (approx. 1 µM; 0.3 µg/mL), and all results were positive and consistent with PM-ECL assay results. However, quantitative detection was difficult with the immunoassay method. On the other hand, the proposed PM-ECL method is expected to be used as a simple on-site analytical method because quantitative analysis can be performed simply by diluting the sample solution. For actual analytical scenes, primary screening by the immunoassay method could be combined with secondary testing by PM-ECL for forensic and clinical testing.

Conclusion

Initially, ECL responses could be observed for codeine in conventional potential sweep mode using [Ru(bpy)₃]²⁺ as an ECL probe under weakly alkaline conditions (pH 9.0). However, methylephedrine and chlorpheniramine, which are found in drugs and pharmaceuticals, also exhibited ECL responses. Morphine, a metabolite of codeine, also showed ECL activity. Therefore, selective detection was difficult. To overcome this issue, we focused on a chemical structure in codeine other than the aliphatic amine moiety, which exhibits ECL activity. Our results from the DFT calculation suggested that the methoxy group at position 3 in the opioid structure of codeine was the electrochemical intermediate because of its energetic stability and steric properties. Consequently, we confirmed that an ECL response for the codeine/[Ru(bpy)₃]²⁺ system could be obtained under weakly acidic conditions with negligible influence from interfering components. However, because the electrochemical oxidation response of codeine was relatively weak, the detection sensitivity by the conventional ECL method was not sufficient. Therefore, we applied the highly sensitive PM-ECL determination technique to the detection of codeine and succeeded in detecting codeine with a sensitivity one order of magnitude higher than that of the conventional ECL measurement. The proposed PM-ECL technique was applied to quantitative analysis of codeine in OTC drugs, and highly sensitivity detection was achieved by simple pH adjustment (pH 5.0) without interference from other components. The proposed measurement system is an innovative method and could be developed into a simple sensor in future by combining the method with a simple mathematical analysis of the measurement results.

Acknowledgments

This work was supported by a Grant-in-Aid for Scientific Research (B) and (C) from the Japan Society for the Promotion of Sciences (JSPS) [Grant Numbers 20K10553, 22H0173, 23K04787, and 23H03177]. Part of this work was supported by Shinshu University Gender Equality Promotion Centre, Researcher Assistance Program.

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

References

• 1) Leppert W., Curr. Pain Headache Rep., 15, 271–279 (2011).
• 2) Yoshimatsu K., Kiuchi F., Shimomura K., Makino Y., Chem. Pharm. Bull., 53, 1446–1450 (2005).
• 3) World Health Organization, “WHO guidelines for the pharmacological and radiotherapeutic management of cancerpain in adults and adolescents,” WHO, Geneva, Switzerland, 2018.
• 4) United Nations, “Narcotic Drugs 2022,” United Nations Office on Drugs and Crime, Vienna, Austria, 2023.
• 5) Kuroyama S., “Drugs with the same indications,” Jiho, Tokyo, 2018.
• 6) Konstantinova S. V., Normann P. T., Arnestad M., Karinen R., Christophersen A. S., Morland J., Forensic Sci. Int., 217, 216–221 (2012).
• 7) Tanaka C., Kato R., “NEW Pharmacology,” Nankodo Co., Ltd., Tokyo, 2007.
• 8) Ahmed S. R., Chand R., Kumar S., Mittal N., Srinivasan S., Rajabzadeh A. R., Trends Analyt. Chem., 131, 116006 (2020).
• 9) Sharma S. P., Purkait B. C., Lahiri S. C., Forensic Sci. Int., 152, 235–240 (2005).
• 10) Maurer H. H., Anal. Bioanal. Chem., 381, 110–118 (2005).
• 11) Takahashi F., Kazui Y., Miyaguchi H., Ohmori T., Tanaka R., Jin J., Sens. Actuators B Chem., 327, 128902 (2021).
• 12) Shiotsuki T., Yoshitome K., Moriya F., Forensic Toxicol., 30, 162–167 (2012).
• 13) Xirui C., Yi L., Wenjing Z., Yuankui L., Yonghua X., Talanta, 186, 29–35 (2018).
• 14) Soni S., Jain U., Burke D. H., Chauhan N., J. Electroanal. Chem. (Lausanne), 910, 116128 (2022).
• 15) Krauss S. T., Remcho T. P., Lipes S. M., Aranda R. 4th, Maynard H. P. 3rd, Shukla N., Li J., Tontarski R. E. Jr., Landers J. P., Anal. Chem., 88, 8689–8697 (2016).
• 16) Thirukumaran M., Singh V., Arao Y., Fujito Y., Nishimura M., Ogura T., Pawliszyn J., Talanta, 231, 122317 (2021).
• 17) Kotani A., Kusu F., Takamura K., Hakamata H., J. Electrochem. Soc., 167, 037517 (2020).
• 18) Afkhami A., Khoshsafar H., Bagheri H., Madrakian T., Sens. Actuators B Chem., 203, 909–918 (2014).
• 19) Afkhami A., Gomar F., Madrakian T., Sens. Actuators B Chem., 233, 263–271 (2016).
• 20) Bahadori H., Majidi M. R., Alipour E., Microchem. J., 161, 105724 (2021).
• 21) Mynttinen E., Wester N., Lilius T., Kalso E., Mikladal B., Varjos I., Sainio S., Jiang H., Kauppinen E. I., Koskinen J., Laurila T., Anal. Chem., 92, 8218–8227 (2020).
• 22) Wester N., Mynttinen E., Etula J., Lilius T., Kalso E., Kauppinen E. I., Laurila T., Koskinen J., ACS Omega, 4, 17726–17734 (2019).
• 23) Miao W. J., Choi J. P., Bard A. J., J. Am. Chem. Soc., 124, 14478–14485 (2002).
• 24) Richter M. M., Chem. Rev., 104, 3003–3036 (2004).
• 25) Brown K., Dennany L., Pure Appl. Chem., 94, 535–545 (2022).
• 26) Dennany L., O’Reilly E. J., Keyes T. E., Forster R. J., Electrochem. Commun., 8, 1588–1594 (2006).
• 27) Dokuzparmak E., Brown K., Dennany L., Analyst, 146, 3336–3345 (2021).
• 28) McLean C., Brown K., Windmill J., Dennany L., J. Electroanal. Chem. (Lausanne), 921, 116649 (2022).
• 29) Mathew E. M., Moorkoth S., Rane P. D., Lewis L., Rao P., Chem. Pharm. Bull., 67, 88–95 (2019).
• 30) Uchikura K., Chem. Pharm. Bull., 51, 1092–1094 (2003).
• 31) Takahashi F., Nitta S., Shimizu R., Jin J., Forensic Toxicol., 36, 185–191 (2018).
• 32) Qiu B., Chen X., Chen H. L., Chen G. N., Luminescence, 22, 189–194 (2007).
• 33) Michel P. E., Fiaccabrino G. C., Rooij N. F., Koudelka-Hep M., Anal. Chim. Acta, 392, 95–103 (1999).
• 34) Jirka G. P., Nieman T. A., Mikrochim. Acta, 113, 339–347 (1994).
• 35) Venkatanarayanan A., O’Connell B. S., Devadoss A., Keyes T. E., Forster R. J., Electrochem. Commun., 13, 396–398 (2011).
• 36) Jin J., Takahashi F., Kaneko T., Nakamura T., Electrochim. Acta, 55, 5532–5537 (2010).
• 37) Takahashi F., Shimizu R., Nakazawa T., Jin J., Ultrason. Sonochem., 63, 104947 (2020).
• 38) Liu Y. M., Tian W., Jia Y. X., Yue H. Y., Electrophoresis, 30, 1406–1411 (2009).
• 39) Pratiwi R., Noviana E., Fauziati R., Carrao D. B., Gandhi F. A., Majid M. A., Saputri F. A., Molecules, 26, 800 (2021).
• 40) Li F., Song J. X., Shan C. S., Gao D. M., Xu X. Y., Niu L., Biosens. Bioelectron., 25, 1408–1413 (2010).
• 41) Li F., Song J. X., Gao D. M., Zhang Q. X., Han D. X., Niu L., Talanta, 79, 845–850 (2009).
• 42) Wu Y. D., Lai D. K. W., J. Org. Chem., 61, 7904–7910 (1996).
• 43) Takahashi F., Nitta S., Shimizu R., Shoji T., Tatsumi H., Jin J., Anal. Chim. Acta, 1191, 339229 (2022).
• 44) Bard A. J., Faulkner L. R., White H. S., “ELECTROCHEMICAL METHODS: Fundamentals and applications,” Wiley, U.S., 2022.
• 45) Takahashi F., Kobayashi K., Jin J., Anal. Bioanal. Chem., 408, 7473–7479 (2016).
• 46) Takahashi F., Matsuda K., Nakazawa T., Mori S., Yoshida M., Shimizu R., Tatsumi H., Jin J., Sep. Sci. Plus, 6, 202200081 (2023).
• 47) Takahashi F., Jin J., Luminescence, 23, 121–125 (2008).
• 48) Takahashi F., Jin J., Anal. Bioanal. Chem., 393, 1669–1675 (2009).
• 49) Takahashi F., Hattori K., Matsuoka M., Jin J., Anal. Sci., 32, 443–447 (2016).